Six Sigma Isn't Continuous Improvement

This month let’s examine another common mistake that some people (not my loyal, intelligent, heretical and, let’s face it, downright attractive readers) make when they use Six Sigma.

I expect to seriously annoy some practitioners when I say that Six Sigma isn’t a method of continuous improvement—no matter how many times you’ve heard someone say exactly that.

Bold assertion, you say? Remember, Six Sigma heretics don’t accept arguments from authority, so keep an open mind and don’t take me at my word. Let’s take a more detailed look.


What’s continuous improvement?

Six Sigma was originally conceived as a way to eliminate defects. If you look at the define-measure-analyze-improve-control (DMAIC) method and history, you’ll see exactly what it does: It closes large gaps in performance or eliminates big problems. Within the Six Sigma context, projects are performed by Black Belts, who are highly trained in advanced tools for solving problems. How many times have we heard the tired phrases “quantum leap” or “breakthrough improvement” when referring to the results of a Black Belt project?

So, if I have these highly trained employees running around solving problems, isn’t that continuous improvement?

I define continuous improvement as “The ongoing actions of taking process inputs at every level of a business, identifying and prioritizing improvement opportunities, and deploying and monitoring local resources against these opportunities, leading to a system that prevents backsliding.” More succinctly, continuous improvement is making all of your processes better through time. The subtlety here is what “better” means.

To paraphrase Dr. Deming, eliminating problems doesn’t improve a process; it allows you to find the process. That underlying variability of the process-as-designed was overshadowed by the problem, and once you eliminate the problem your process goes back to its default state. It isn’t any better than it was—you just stopped messing it up. Once you’ve found the process, you can start to improve it by reducing that underlying variability.

So, sure, it’s better for your business to have the problems solved, but the underlying process is the same—it hasn’t improved.

How would you go about improving your processes, then? Well, you should first turn to the people who are constantly in contact with that process—“process owners.” Everyone at your company is a process owner. Yet, if we look at how many Black Belts most companies have, it’s maybe 2 to 10 percent of their personnel. So what are the other 90 to 98 percent of these employees doing?

Well, um… yeah, running their process. And what else?

Part of what they should be doing is taking some time to improve their process. I mean, who better knows the annoying characteristics of the process status quo than the people who use it every day? Top management needs to provide them with a method, time and resources for them even to consider improving the process, but whatever this activity amounts to, it sure isn’t Six Sigma. Nonetheless, we’ll see that this process complements Six Sigma and can take advantage of some of the same basic tools. The key is to enable all of your employees to improve their own process on a never-ending basis. These probably aren’t big breakthrough changes and only rarely call for the advanced tools that Black Belts learn. The Japanese call this kaizen. I call it necessary to survive in a global economy.

It gets worse. Continuously solving problems as Black Belts do sounds great, even exciting. But where’s the biggest return on your investment—fixing problems after they become large enough to warrant bringing in a Black Belt, or fixing them before they get that big? People are really good at following the path of reward, monetary and otherwise. If you make it clear (e.g. stock options, recognition, etc.) that you want your big problems to get solved by a cadre of the best and brightest, you’ll have a never-ending supply of big problems, not because people are creating them on purpose, but because you aren’t supporting problem solving locally when the problems are still small. You must make it clear that, while solving problems is appreciated, you want your employees to be able to prevent problems from even occurring. This drives a set of behaviors that Black Belts are fundamentally unable to perform as Black Belts, because that isn’t their job function.

Continuous improvement isn’t continuous problem solving

I contend that continuous problem solving is quite different from continuous improvement. If you’re continuously solving problems, you probably have a system that is continuously degrading to produce those problems, probably because that continuous improvement is seen as the job of the Black Belts. The Red Queen in Lewis Carroll’s Alice in Wonderland may describe your business when she says, “Now, here, you see, it takes all the running you can do to keep in the same place. If you want to get somewhere else, you must run at least twice as fast as that!” This is one way you can have Black Belts claim millions of dollars in savings and yet not see it in the bottom line. Instead, you should be using some of your resources—the other 98 percent of your workforce—to maintain and gradually improve their own processes.

Therefore, Six Sigma isn’t continuous improvement. Black Belts are trained to attack a few big-ticket problems, and there are too few of them to work on all of your processes all the time anyway. Continuous improvement is done by everyone else on their own processes, while your Black Belts are working on a few strategic breakthroughs.

I consulted for a vice president of quality who phrased it like this. “I can see how Black Belts are like a rifle you take with you to the woods—they are great for shooting big problems such as bears. The problem is I am up to my waist in snakes!”

There are two functions that need to be enabled with supporting infrastructure. The first is continuous improvement using lots of people to make lots of ongoing, small improvements in the processes they know and love, capitalizing on their process knowledge and the benefit they’ll realize on their own job. The second is breakthrough using a few highly-trained and focused individuals to work on solving a few critical problems that will have big effects on the business.

Does this take us Black Belts down a notch? No, and even if it did, is that really such a bad thing? The work Black Belts do is essential in the near term for most companies in order to stop the arterial bleeding of red ink. But if they do their job right and big problems stay fixed, and if we also have ongoing continuous improvement—where little problems are permanently fixed before they become big problems—then deploying a Black Belt to solve a problem should be viewed as a failure of the system, not as a potential success.

Heresy indeed.

The consequences of Stupid Six Sigma Trick # 9

Imagine what would happen if someone naïvely thought that Six Sigma and Black Belts were their chosen method of continuous improvement. Because we’re assuming that continuous problem solving is continuous improvement, we expect to see big bottom line gains. Our loyal Black Belts work hard and complete $10 million worth of savings by working on 5 percent of the processes in the business, but, in the meantime, the other 95 percent of the processes have degraded a small amount each, which by chance adds up to $10 million.

At the end of the fiscal year, we write a nice letter to the shareholders telling them about our savings due to Six Sigma. Then we get a call from our shareholders, asking that if we saved $10 million, could they please have their share. Now. Holy cow! No matter where you look, there isn’t that much more money to hand out. So your shareholders think that either you’re incompetent or that you have a secret bank account in the Cayman Islands.

Meanwhile, your front-line people see that, while you’re proclaiming victory, most of their processes are getting worse with this Six Sigma thing and will be happy to tell anyone (at length) just that.

Final thoughts

To avoid stupid Six Sigma trick #9, you need to understand clearly that there’s a difference between continuous improvement (many people making small improvements in many processes on an ongoing basis) and breakthrough (a few people making a few big improvements in a few processes) and that you need to create an infrastructure to support both activities. Many tools can be used for both functions, and Black Belts can be great resources for your continuous improvement teams. The good news is that while your Black Belts are working at solving big problems, the rest of the company is busy working on their local problems. Interestingly, the infrastructure to support the local continuous improvement work is also what monitors the efficacy of the changes the Black Belts make once they complete their project and hand it back to the process owners. These two functions aren’t either/or—both are necessary and interlocking.

Then again, I could be wrong.